FIG. 3 is a diagram illustrating a cross section of a prior art solid-state imaging device disclosed in the International Electron Devices Meeting, 1982, p.324. In FIG. 3, reference numeral 1 designates an n type Si substrate. A first p conductivity type region 2 is produced in the Si substrate 1. A first n conductivity type region 4 which is to be a channel of a charge coupled device (hereinafter referred to as CCD), a second n conductivity type region 5 carrying out the photo detection, and a second p conductivity type region 3 serving as a channel stopper are produced in the first p conductivity type region 2. A third p conductivity type region 6 is produced in the second n conductivity type region 5 to suppress the generation of dark current at the interface between the Si substrate 1 and an SiO.sub.2 film 7. A charge transfer electrode 8 of the CCD comprises a polysilicon film formed on the surface of the element.
FIG. 4 is a diagram showing the impurity concentration distribution taken along the line B--B' of FIG. 3. As shown in FIG. 4, impurity concentration distributions are formed in the order of p, n, p, and n type from the surface of the element.
A description is given of the operation.
When light is incident to the element from the above, photoelectrons are generated in the depletion layers adjacent to the pn junctions which are produced between the second n conductivity type region 5 and the first and third p conductivity type regions 2 and 6 and the photoelectrons are stored in the second n conductivity type region 5. When a high level voltage is applied to the polysilicon film 8 serving as a charge transfer electrode after a predetermined storage time, the photoelectrons are transferred to the channel 4 of the CCD and further output to the outside.
FIG. 7 is a diagram showing an electric potential distribution taken along the line B--B' of FIG. 3. The function of the third p conductivity type region 6 is described in detail using this figure.
The second n conductivity type region 5 is intended to be completely depleted at the time when photoelectrons are transferred to the channel 4 of the CCD. Otherwise, remaining electrons result an afterimage through the thermal diffusion process. However, when there is no third p conductivity type region 6, complete depletion of the second n conductivity type region 5 means that the depletion extends to the interface with the SiO.sub.2 film 7. There is a defect 16 which produces an energy level in the energy band gap of the silicon at the SiO.sub.2 interface and it functions as a generating center of dark current charge carriers. Accordingly, the third p conductivity type region 6 is provided to produce accumulated holes, thereby suppressing the generation of the dark current caused by the defect at the interface.
A description is given of the production method.
FIGS. 5(a) to 5(c) are cross-sectional diagrams illustrating a production process of the solid-state imaging device having the above-described construction.
Firstly, boron is implanted into the n type Si substrate 1 to produce the first p conductivity type region 2. Phosphorus and boron are selectively implanted through a resist mask and then they are diffused respectively through annealing process to produce the second p conductivity type region 3, the first n conductivity type region 4, and the second n conductivity type region 5 (FIG. 5(a)).
Next, the SiO.sub.2 film 7 is formed on the surface of the Si substrate 1 by a thermal oxidation method. The polysilicon film 8 is deposited using the chemical vapor deposition (CVD) method and thereafter the polysilicon film 8 is patterned into a desired configuration (FIG. 5(b)).
Next, boron 10 is implanted into the entire surface of the element at an energy so that the boron cannot transit the polysilicon film 8 but can transit the SiO.sub.2 film 7, and then the boron 10 is activated by annealing to produce the third p conductivity type region 6.
In the above-described processes, the boron 10 for producing the third p conductivity type region 6 is implanted after patterning the polysilicon film 8 which is to be a charge transfer electrode of the CCD, and thereby the spreading of boron having a large diffusion coefficient is suppressed. If boron spreads to such an extent that the third p conductivity type region 6 extends to the deep portion of the n conductivity type region 5, the effective impurity concentration of the n conductivity type region 5 which stores photoelectrons will be reduced and the storage capacitance is lowered.
The prior art solid-state imaging device constituted as described above has typically two problems.
First, although the third p conductivity type region 6 is produced at a later process so that it spreads as little as possible, it is not possible to avoid annealing. For example, a reflow treatment for producing a flattening film on the polysilicon film 8 is carried out at high temperature or about 900.degree. C. Therefore, the boron ions implanted in the process step of FIG. 5(c) are further diffused during the later process, thereby reducing the effective impurity concentration of the second n conductivity type region 5 which determines the total storage charge amount of the photodiode. In addition, since the pn junction is deep relative to the surface, the sensitivity to blue light which is absorbed near the surface region is lowered.
Second, the electron potential .phi..sub.PD and the maximum storage charge amount Q.sub.PD when the second n conductivity type region 5 is depleted, shown in FIG. 7 vary as shown in FIG. 8 in response to the impurity concentration of the n conductivity type region 5. While the maximum storage charge amount Q.sub.PD needs to be more than a predetermined value (shown by an upward directed arrow in FIG. 8), the electron potential .phi..sub.PD needs to be less than a potential value determined by the voltage applied to the charge transfer electrode of the CCD (shown by a downward directed arrow in FIG. 8). The relation between the maximum storage charge amount Q.sub.PD and the electron potential .phi..sub.PD can be represented by the formula Q.sub.PD =C.sub.PD .times..phi..sub.PD, using the capacitance C.sub.PD of the depletion layer adjacent the pn junction. Accordingly, in order to satisfy the above-described two conditions, the maximum storage charge amount Q.sub.PD can be increased by increasing the capacitance C.sub.PD of the depletion layer, that is, by narrowing the depletion layer at both sides of the second n conductivity type region 5 even keeping the same electron potential .phi..sub.PD. However, incident light is converted to charge carriers in the depletion layer and the number of generated electrons as a function of the incident light amount depends on the width of the depletion layer. Therefore, when the depletion layer extending at both sides of the second n conductivity type region 5 is narrowed to increase the capacitance C.sub.PD of the depletion layer, the sensitivity of the element is reduced.
As described above, in the prior art solid-state imaging device, it is difficult to form a shallow and high-impurity concentration p conductivity type region 6 that reduces the generation of dark current at the surface of the photodiode. It is also difficult to satisfy the desired conditions for the electron potential and for the maximum storage charge amount required to maintain good sensitivity.